Modeller, araçlar ve bilgi erişimiyle LLM uygulamaları ve ajanlar oluşturmak için açık kaynaklı çerçeve
AI platformları AI geliştirici platformları docs.langchain.com
What is LangChain
LangChain is an open-source framework for integrating large language models into applications, used to build chatbots, retrieval-augmented generation (RAG), document analysis and AI agents. Harrison Chase launched it as an open-source project in October 2022, and it later became a company of the same name. LangChain 1.0 was released on 22 October 2025, introducing a create_agent abstraction and middleware for customising agent behaviour.
What it can do
- One interface for many models: connect to models from different providers through the same API, with fewer code changes when you switch
- Build agents: LangChain is the quick way to build an agent, and LangGraph, from the same team, gives finer control over the flow
- RAG and documents: connect vector databases and document loaders for knowledge-base Q&A and summaries
- LangSmith: the company's closed-source platform for tracing agent runs, evaluating results and deploying to production
Pricing
The LangChain framework is free and open source, with its GitHub repository under the MIT licence and about 150,000 stars. Hosted platforms such as LangSmith are paid; see the official website for plans.
Who it's for
- Engineers building AI applications and agents in Python
- Teams that connect several model providers or build RAG knowledge bases
- Students and developers learning to build LLM applications
Good to know
- Version 1.0 narrowed the main package and moved legacy features to langchain-classic, so code from older tutorials may need changes.
- For simple calls, a model provider's own SDK can be lighter; reach for a framework when the flow gets complex.
Similar tools
Compare LlamaIndex, Dify and Coze; for low-code agents, also see n8n.